Creating Enlarged Separation Negatives - How best to do it?

[holmburgers]

As I go forward making dye-transfer prints via DCG-matrices I'm not going to be content with 4x5" prints for very long.

I need to consider my options for making enlarged separation negatives at affordable prices. What makes this trickier than negs for monochromatic prints is the need for a pan-sensitive separation material. But buying 11x14"+ panchro sheet film is not a sustainable option. ($)

A less expensive emulsion on a clear substrate is what I'm looking for. This will undoubtedly entail ortho, so I'll need to transfer the original color image 1:1 to a panchromatic b&w film. Then this can be enlarged. Depending on whether I'm coming from an E6 slide or a C41 neg, this will require a different procedure.

Broadly, what is there that's inexpensive, available in large sizes and on a clear substrate?

What is the current state of lith film? I'm not easily finding it for sale. Maco Genius Print Film is such a material, but I don't think it's available in the US. I found Efke Print Film but that is quite expensive.

Working out the procedures is one thing, but at the moment I'm not sure if there's an available material that's ideal.

Help!

I'm curious about x-ray film which I understand can be had for cheap. Another crazy option would be to pour liquid emulsion on Photoformulary melinex.

Depending on what film I find, matching the contrast and gamma characteristics of potentially 5-6 generations is going to be a feat... gulp.

But where little is ventured, little is gained!

Furthermore, I'm not interested in digital negatives at the moment as any means I have to produce them would be completely inferior to analog methods. Though, I admit that its convenience is enviable.

 

[michaelbsc]

I have been thinking about this too.




Recently I bought an old slide duplicator, which is essentially a 35mm camera body with a bellows in front of the lens set up like a copy stand.




My thought is to change the light source to get the RGB separation.





I think this could be scaled up larger with a little work.





Pan film in the camera can then be enlarged onto other mediums.

 

[holmburgers]

I think that's what I'd do too and the recommendation of changing the light source is a good one.

But what to enlarge to?? That's what's got me stumped.

If anyone can find a link to lith film to purchase, please show me! I can't tell if I'm going crazy or this stuff has dropped off the earth... that, or I'm stuck in an internet blackhole.

Any plans for your separation negs? Triple projection tri-color or something??

 

[michaelbsc]

I haven't paid attention lately. Too many things in life going on.

I know Freestyle carried lithograph film, and Ultrafineonline used to have it too.

I'll be very disappointed if it's gone forever.

 

[holmburgers]

This seems promising... http://www.ultrafineonline.com/ulcotodufi.html

And what the heck is this stuff? -> http://www.ultrafineonline.com/rlronefi.html & http://www.ultrafineonline.com/rlrodufi.html

 

[holmburgers]

And this... can it be tamed for continuous tone?

http://www.ultrafineonline.com/ulhicoorlifi.html

 

[Michel Hardy-Vallée]

It's just my reading, and pardon me if I am skipping over an important detail of DT printing, but if you follow Michael's procedure and use a slide duplicator, then your 35mm copy film should be panchro, not lith or ortho.

If that's the case, then you can go with TMX 100 which can be processed to either a very low or a very high contrast depending on your needs.

 

[holmburgers]

Hi Michel,

Yes, that's of course part of the plan, to copy to panchro and then ideally enlarge to ortho from a monochrome. I just don't know what the best material for enlarging is. I'm looking for something that's relatively affordable in large sizes and has suitable characteristics for dichromated-gelatin.

 

[greybeard]

holmburgers,

If I understand correctly, you will do your tricolor separation at the first duplication (slide copier or whatever), and then enlarge the three monochromes. For image quality, you should probably enlarge as much as practical in the color separation step (i.e,, 35mm or 6x6 color to 4x5 tricolors), then contact print the positives to get the negatives which you will enlarge to final size. This will allow you to work out the various transfer functions on relatively small film, and only require panchromatic stock for the first step. Either ortho or blue-sensitive film would work for the last two steps; the choice would be driven by both cost and tractability of the H&D curves. (Doing this from a C41 original sounds like a nightmare to me....)

Or am I missing something here?

 

[holmburgers]

Thanks Greybeard,

That's the basic idea, but part of the problem is the cost of larger panchromatic materials, therefore I'd like to go 1:1 in the separation phase, with a high resolution film. Ideally contact printing, but especially with 35mm that sounds like a pain in the arse. Then, enlarge to a large cheap film.

To be honest, I'm exploring the digital route, as much as it pains me to say it. But I'm not giving up ultimately, I just think that for testing it might be easier.

We'll see, I'm stil not convinced that there's a suitable material (that's "cheap")

 

[greybeard]

That's what I thought you had in mind, but it seems that the cost of three sheets of 4x5 pan film for the separation negatives would sort of disappear into the overall cost of working out the transfer functions (unless you are planning to do an awful lot of these pictures). I would imagine that the best argument for doing the separation at 1:1 is that the color correction of the enlarging lens would no longer matter; the best argument against it is that when you enlarge, you will be enlarging any defects present in both the original and the contact print by the same amount.

In other words, a direct 35mm-to-11x14 is an enlargement of about 11X, and any dust, scratch or whatever on the contact will be enlarged by 11x, along with anything on the original. But if you go to 4x5 first, this is only about 4x, so anything you pick up at the 4x5 stage (where you would also do your reversal, if necessary) will be enlarged by 2.75X, and your tweaking of the curves can be done on 4x5 film for consistency. Test strips from the 11x14 stock will require yet another process adjustment, as if there weren't enough already

There is a reason why the printing industry abandoned tricolor separation negatives as soon as digital processing became practical....

It just seems that with panchromatic 4x5 films running $.50 to $1.00 per sheet, it would be hard to break the bank at the color separation step; x-ray film (either blue-sensitive or ortho) would then work for the final enlargment.

 

[holmburgers]

I definitely see your point about dust and stuff. Not too mention, it would be nice to standardize on one emulsion, one processing procedure. That would leave me to figure out enlarging to film and contact printing to film; doable I s'pose.

How would one go about figuring out the 'transfer functions'? What kind of testing would be the most efficient use of materials? I guess I would need to find out a reliable method for knowing the density of the negative, and then how that relates to exposure, either by contact or enlargement, and then figuring out how development plays in. All in all, a tall order for someone with zero densitometry experience. But, I'd like to learn it.

 

[greybeard]

I suspect that the most valuable resource at this point would be the library of a university or college with a robust graphic arts department. If you are only doing monochrome, then life is good; figure out the required density range of the negative in order to get the full range of the final material, and you are all set.

Color is quite a bit more complicated, since the (minimum) three colors will in general not be "pure" and you will be trying to match both the overall hue (color) and saturation (density). When dye-transfer materials were in vogue, there was probably a lot of data available, but by now this is probably in only a few libraries or the hands of the remaining practitioners. Color gum printing is still done, so there may be some tutorial information available from that field.

In the absence of a color densitometer (and possibly in the presence of one!) you can always invoke artistic license to cover any departures from colorimetric correctness. After all, hand-colored prints are rarely very realistic, but they are often quite beautiful anyway.

I'd love to see where all this leads you!

 

[Michel Hardy-Vallée]

Let's think about the procedure...

If the original is a colour positive:

Positive (slide) -> 3x B&W internegative (TMX 100) -> 3x dye transfer matrix

You can make the internegative by contact (get a Leitz ELDIA!) or with a slide copier. Then you enlarge each of your separation internegatives on the matrix film. No expensive pan sheet film needed.

If the original is a colour negative, you have either:

1) Negative (C41) -> 3x dye transfer matrix (with pan matrix film)
2) Negative (C41) -> 3x B&W internegative (TMX 100) -> 3x interpositive (copy film) -> 3x dye transfer matrix (non-pan)

I know Ctein has Pan matrix film left, but he may be the only one... So you're left with solution 2).

In solution 2), you can do the same procedure as above with an ELDIA or a slide copier. Then you enlarge your internegative onto copy film. Efke print film, for example, is a slow film designed for this kind of use. Have a look at the Specialty Film section of Black & White films on Freestyle to choose one. Once your enlarged interpositive is done, you contact print it onto matrix film.

Yet again, no expensive pan film needed. The real problem you're left with is the question of contrast. The gamma, or slope of the straight part of the H&D curve evolves from generation to generation via a multiplication operation. Demonstration:

Let's say for example that you have an original negative of average contrast (gamma=0.6). If you copy it onto TMAX film, and develop TMAX as usual (gamma = 0.6), the resulting image will be way too low (0.6 * 0.6 = 0.36 !!). So that's why when you make B&W slides out of a normal negative you need to boost the hell out of the copy film's contrast to attain something decent.

A slide has a higher gamma than a negative, around 1.2 or 1.5. If you want a final image with a gamma of 1.2, and your original negative has a gamma of 0.6, then your copy film must be developed to a gamma of (x) in the following equation: 0.6 * (x) = 1.2. If we solve the equation, the answer is 2. Getting a film to reach a gamma of 2 requires usually ortho film in stock Dektol or something like that...

Now, if we apply this logic to your situation, you need to work backward. What's the gamma needed by your matrix film? Once you know this, you can start solving the (x) for your internegatives/interpositives. Your typical colour neg has 0.6 gamma, your typical slide a 1.2 gamma.

There are obviously a lot of other issues that I'm skipping over by omission or ignorance, but I thought you could use a short primer on gamma. It helped me a lot to figure things out when I made B&W slides. At least I knew I had to develop my slides A LOT so that they looked good!

 

[holmburgers]

I appreciate the primer on gamma. Presumably my desired gamma will be similar or at least related to the gamma needed in carbon printing.

See, I'm not going to be using Kodak matrix film. A dye-transfer (imbibition) matrix is nothing but an image comprised of hardened & unhardened gelatin. Kodak's dye-transfer matrix film used silver-halides and a tanning developer, but since the materials no longer exist (for all intents and purposes), this is a much harder thing to do.. to make an emulsion.

Instead, I'm going to use the carbon principle to form my hard/unhardened images in gelatin, and dye is "imbibed" (absorbed) according to the thickness of the relief image. So I need negatives the size of my print.

 

[greybeard]

It was my understanding that you were planning to make dye-transfer matrices using dichromated gelatin or the like; this would be exactly analogous to the old Pan Matrix Film, except for the chemistry of the tanning/hardening and possibly the use of a water development step. In either case, each matrix is the size of the finished print (to within a few thousandths of an inch, if you want fine detail!)

Now, if you do not plan to transfer the dye from the matrix to a permanent support, it is hard to see how you will be getting tricolor reproduction. Three or four dyed gelatin images on independent supports, superimposed in register? This would give you a color transparency, but it is hard for me to see how you would make a color image on an opaque support---three gelatin layers would work, as in tricolor gum, but you will need dyes (as opposed to pigments, which are insoluble and pretty inert) that will stand up to UV light while saturated with dichromate and not leach back out of the gelatin in successive steps. (Dye transfer depends explicitly on transfer by diffusion from the matrix to the support.)

I still think that however you proceed, you will need a systematic way to manage the density (saturation) of the dye images as well as the contrast; this is a major problem in color printing press management (another one is registration...) so perhaps a good graphic arts textbook would be the place to start.

 

[holmburgers]

Yes, I'm making dye-transfer matrices which will be like pan/ortho-matrix film in many ways except that they must be contact printed, since enlarging by UV onto a dichromated-gelatin matrix isn't really possible (though the Fresson process does.... hour+ exposures with carbon arc!)

This is the reason that Kodak's dye-transfer scheme was so commercially successful; the ability to enlarge from small negatives. The same can be said for carbro printing.

Now a pan-matrix film... well that's just sheer luxury compared to what I'm proposing.

All this being said, of course I plan to transfer them to an opaque paper support.

After you hot-water-etch the DCG-matrix (exactly like Kodak's DT process) you are left with a positive-relief matrix, with varying depths of gelatin that dye will be imbibed to, in proportion to its thickness. You control contrast of the dye during the transfer phase with acid/buffer concentrations in your rinse. Getting a good matrix by using the carbon method will be necessary to fully explore the controls available with acid/buffer concentrations.

The problem at the moment is not the mechanics of a dye-transfer system, on which I've done a lot of reading, but simply the best way to make enlarged negatives.

 

[greybeard]

Ah. It was the contrast/density of the dye image that had me baffled. I didn't know that the acid and buffer concentrations would give you another "knob". Presumably, if you are using the same dyes and similar gelatin to the Kodak process, the colorimetry will take care of itself.

In that case, the "target" density and contrast are probably documented, and Michel's primer is right on. The only other thing could be contrast masking, but this should also be covered in dye-transfer technique.

After giving it all some thought, I still can't come up with anything better than doing the separation in the first step, and then enlarging the monochrome images onto x-ray film. Once you get larger than 8x10, there are essentially only two product classes to choose between: x-ray film (cheap, but often double-coated, no antihalation, and designed for very high Dmax) and photographic film (expensive). In that light, the relative merits of single- versus two-stage enlargement are pretty much academic.

I suppose that you could also consider a hybrid approach to get started: use a service bureau for color separation negatives at full size of one or a few prized originals, then work out the "back end" processes that involve personal judgement and expression. When you have the gum, dye, chemistry and timing down, revisit the "front end" steps to replace the (d*****l) service bureau with a purely analog technique.

For my own interest, I just looked up ortho and pan sheet films at B&H, and was surprised to see that 4x5 ortho sheet is right at 6 cents per square inch, 8x10 ortho is a good bit cheaper at 4.7 cents, and 400-speed pan film in 11x14 is even better, at 4.2 cents per square inch. But if you can figure out how to use Agfa white-light x-ray duplicating film, you can get it for about half a cent per square inch!

 

[holmburgers]

As I understand it (having never done it yet) dye-transfer is a very controllable process, but matching up all the variables certainly seems like quite an enormous task.... too many knobs!

So.. x-ray film it might be then. You're definitely making it sound good at .5¢/in². That's amazing, thanks for the numbers.

I remember a couple of really good threads about x-ray film, one of which I can't seem to find. I'll keep digging though..

Also, I agree with your comments about getting sep. negs. made in some capacity. Once the kinks are worked out, the expense of LF pan or ortho film might be justified.

 

[greybeard]

holmburgers,

I guess I'm just having one of those days; I can't seem to see how the cost of pan film could be that big an obstacle.

Without allowing for shipping and/or taxes, using B&H prices for FP4+ and CSX for x-ray duplicating film, enlarging a color negative through filters onto three sheets of 4x5 FP4+ would cost you 3 x $0.86 - $2.58 for pan stock. Enlarging these tricolor positives onto 14x17 x-ray film would then cost you no more than 3 x $0.65= $1.95, so the marginal cost for a set of tricolor negatives would be $4.53. (I wish I could always get a presentation-quality 8x10 print without its costing more than that in paper!)

If the CSX "white light" duplicating film is in fact panchromatic and has decent curve shape, then you are basically home free (or, rather, home for only $1.95 ).

 

[holmburgers]

I mean that 11x14" pan film would be an obstacle. Yes, 4x5" is fine and I have a lot of it.

Where are you finding the x-ray dupe film? I'm searching for CSX but haven't come up with anything yet....

I recall a discussion about the white light and I think it is not panchromatic unfortunately.

Greybeard, this is great... you're helping me put this puzzle together!

 

[greybeard]

Try here: http://www.cxsonline.com/text/detai...LQG&cart=130142992461657434&location=10011003

I notice that I quoted the price for 10x12 duplicating film, but mentioned 14x17 (it's a whopping $1.27 per sheet...)

There is an extensive thread on x-ray film over on the LF forum, but it is nearly all about using the film in-camera. At some point, I'm just going to have to get some to play with myself.

 

[holmburgers]

Solid!

Here's a pertinent thread, haven't even read through it myself though. (there was a url link here which no longer exists)

 

[holmburgers]

http://www.cxsonline.com/text/detai...LQG&cart=130142992461657434&location=10011003

http://www.mackiexray.com/

This might really be the ticket. It's so cheap and it's super high contrast, which I think might be a plus with a carbon-system. And apparently they send out free samples....

 

[greybeard]

Wow. It's been a while since I've read anything that detailed on color separation. I sort of knew what was involved, but it's a shock to see it confirmed....

Thanks, Greg!